CUDA/HIP: refractor mmqv to unify the calculation of nwarps and rows per block between host and device code. #12177
+142
−59
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
github-actions
bot
added
Nvidia GPU
…between host and device code.
…s in device code, even though that should not be a problem.
JohannesGaessler
approved these changes
Mar 7, 2025
Co-authored-by: Johannes Gäßler <[email protected]>
ishaangandhi
pushed a commit
to ishaangandhi/llama.cpp
that referenced
this pull request
Mar 12, 2025
…per block between host and device code. (ggml-org#12177) refactor mmqv to unify the calculation of nwarps and rows per block between host and device code. --------- Co-authored-by: Johannes Gäßler <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This refactors mmqv to unify the handling of parameters between host and device side code, avoideing duplication in calculateing nwarps and rows_per_cuda_block. Also explicitly handles wave_size != 32, for the minor benefit of getting us out of shared memory into warp level primitives one iteration earlier.